PhD Defense : Lisa Altinier
22 April à 14h00 - 17h00
“Development of high-order mode dithering: a new image processing method for exoplanet detection with the NASA Roman Space Telescope”.
The defence will take place on Wednesday, 22nd April at 2 p.m. in the LAM amphitheater. Please find the abstract and the jury composition below. The defence will be in english.
– Élodie Choquet (LAM, CNRS, Marseille) – Thesis co-director
Abstract:
Direct imaging of exoplanets is crucial for characterizing their physical and atmospheric properties, but is challenging due to high contrast with host stars and small angular separations. Indirect methods like radial velocity and transits detect many exoplanets but provide limited atmospheric information. Achieving contrasts of ∼ 10−9 for Jupiter-like planets or ∼ 10−10 for Earth-like planets in reflected light exceeds current telescope capabilities. Ground-based adaptive optics detect young, infrared-bright planets, while Hubble and JWST lack active wavefront control for high contrast coronagraphy.
The Nancy Grace Roman Space Telescope (Roman, RST), launching in 2026, addresses this gap with a 2.4 m telescope and the first space-based coronagraph with active wavefront control. Its coronagraph can create a dark hole suppressing starlight to ∼ 10−9 after post-processing, enabling detection of old Jupiter-like planets at 3– 9 λ/D (0.15–0.44 arcsec). Quasi-static speckles from optical imperfections, thermal drifts, and DM actuator noise dominate the noise and require post-processing methods such as principle components analysis (PCA) and reference differential imaging (RDI).
Classical RDI is limited by wavefront drift and low reference star diversity. High-Order Mode Dithering (HOMD) injects controlled high-order aberrations (Z5–Z24) into deformable mirrors during reference star observations, enriching the speckle library and improving post-processing. CAPyBARA, a dedicated simulator, models the coronagraph and observing sequence, including static and quasi-static aberrations.
Results show HOMD effectiveness depends on drift, mode selection, and alignment with dominant aberrations. Low-order modes (Z5–Z8) provide the largest improvements, while partial overlap reduces effectiveness. HOMD increases library diversity, mitigates drift sensitivity, and improves post-processed contrast, providing a practical method to enhance space-based high-contrast imaging and paving the way for future missions like the Habitable Worlds Observatory, which will require extreme contrasts (∼ 10−10) for Earth-like planets detection.
Footer
